The PHERGain study showed that a response-adapted approach guided by [18F]FDG-PET imaging, focused on achieving pathological complete response (pCR), allowed safe omission of chemotherapy in select individuals with HER2-positive early-stage breast cancer treated with neoadjuvant dual HER2-targeted therapy (trastuzumab plus pertuzumab, or HP). Given the restricted access to [18F]FDG-PET in many centers, this investigation explored the potential of breast MRI as a substitute modality for monitoring early therapeutic response. In cohort B (n=285), participants started with two cycles of HP alone, with chemotherapy added only for those lacking response on [18F]FDG-PET. Imaging with both [18F]FDG-PET and MRI was performed at baseline (prior to treatment assignment) and after the initial two cycles (early evaluation). A follow-up MRI was also obtained prior to operative intervention (late evaluation). The analysis examined agreement between [18F]FDG-PET findings, MRI-measured tumor diminution, and RECIST 1.1 criteria on MRI, as well as their predictive value for pCR and 3-year invasive disease-free survival (iDFS).
Early assessments revealed strong agreement (78.2% accuracy) between [18F]FDG-PET response and any degree of tumor shrinkage on breast MRI, though concordance was lower when using strict RECIST 1.1 definitions (at least 30% reduction in target lesion diameters). Patients responding on [18F]FDG-PET who also demonstrated early MRI shrinkage or RECIST response had elevated pCR rates (39.0% vs. 29.6% without shrinkage; 44.0% vs. 30.4% without RECIST response). Conversely, [18F]FDG-PET non-responders lacking MRI shrinkage exhibited the poorest outcomes, with pCR of only 21.7% and 3-year iDFS of 75.3%, even after chemotherapy escalation. In [18F]FDG-PET responders continuing HP without chemotherapy, prolongation of therapy boosted late MRI complete responses (from 9.3% to 31.7%) and overall response rates (from 55.1% to 70.0%). Late MRI complete response was more reliable for forecasting pCR in hormone receptor-negative disease (positive predictive value 85.5%) compared to hormone receptor-positive cases (61.5%). While [18F]FDG-PET remains the preferred tool for directing adaptive treatment decisions in the PHERGain framework for HER2-positive early breast cancer, this exploratory analysis indicates that measurable tumor reduction via breast MRI may serve as an effective substitute in environments lacking PET access.
